基于归一化编辑距离的自由文本击键特征分类识别方法  

Keystroke Characteristics Classifi cation and Recognition Method of Free Text based on the Normalized Levenshtein Distance

作　　者：王博[1] 胡晓勤[1] 

机构地区：[1]四川大学计算机学院,四川成都610065

出　　处：《计算机安全》2014年第10期15-21,共7页Network & Computer Security

摘　　要：针对基于相对距离的自由文本击键特征分类识别方法 (R方法 ),提出了一种基于归一化编辑距离的自由文本击键特征分类识别方法。通过数据收集、预处理、距离计算、归一化处理、分类判定等过程,实现了对中文自由文本击键特征样本进行准确分类识别的方法。实验证明,本方法比R方法的分类正确率更高,识别准确度更好。According to the keystroke characteristics classification and recognition method of free text based on a relative distance, also named R method, this paper proposes a keystroke characteristics classification and recognition method of free text based on the Normalized Levenshtein Distance. Through data collection, pretreatment, distance calculation, normalized processing, classification judgement and other processes, it implements this method accurately and exactly on Chinese free text. Experiments proved that this method is better than R method on classification accuracy and recognition rate.

关 键 词：相对距离 归一化编辑距离 自由文本 击键 特征识别 分类 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]